DIY DC power cables

[Ryelands]

1 hour ago, E-Stat said:

. . . when I reported my findings over at AA, a self-professed expert said there was no "objective basis" behind the change.

 As it goes, I'd just drafted a comment for that thread when the moderators (understandably) deleted it. To show you weren't shouting into the wind, here's what I would have said:

 

++++

Your claim has no objective basis.

 

I can't speak for E-Stat but the notion that his claim has no objective basis is, to put it politely, ill-informed. As the article linked below explains, quad-core geometry was developed for the telephony sector about eighty years ago for reasons that are easily understood by those who bother to look. Quad-core cables of the type John uses in his "recipe" have been commonplace in pro-audio circles for some fifty years. Modern types are manufactured in bulk by, inter alia, Canare, Supra, van Damme and Mogami. Still, what do they know?

 

That doesn't, of course, automatically mean that they are effective in the application John was discussing but it was easy to confirm that they are. It was clever of JS to spot the potential for an inexpensive industry-standard product to provide it as low inductance is not generally considered critical for microphone cables though, of course, it is for speaker cables (see e.g. Supra data).

 

http://www.soundonsound.com/sound-advice/q-whats-special-about-star-quad-cable

[Ryelands]

3 hours ago, JohnSwenson said:

There are a couple places on ebay UK that are selling short lengths of Canare 4S6.

 

A popular UK equivalent is Van Damme "XKE Star Quad Microphone cable Pro balanced"; I bought mine from "Custom-Lynx" in Leicestershire; Though I can't find the star quad line on the company's web site, it's easily found on ebay. In a choice of ten colours, even. I've tried both the Van Damme and the Canare: If there's a difference between them, I can't spot it.

 

In passing, pinching John's "isolated screen" recipe, I made up a pair of interconnects using Van Damme SQ cable and, while simultaneously swallowing scepticism and gulping at the price, KL Eichmann phono plugs.

 

Though I've never before felt able to use the cliché "not subtle" to describe the difference made by a pair of cables, I don't think the demure shade of white had anything to do with it . . . In a word, excellent. Many thanks!

[Ryelands]

1 hour ago, JohnSwenson said:

. . . an external end to end wire will work well with Ethernet. You can get the advantage of shielding without having to connect the shield to the boxes possibly creating a leakage loop.

 

Indeed it does. I'd found a while back that ViaBlue CAT7 "audio" grade screened LAN cable sounded, despite its lively price, poorer in my setup than a decent unscreened LAN cable such as Excel though the difference was much reduced after I disconnected the screen on the ViaBlue and cut the pairs that 100-BaseT doesn't use (4&5, 7&8).

 

On a whim, I recently restored the ViaBlue after joining up the ends as John suggests. It now seems to sound a little bit better than the Excel cable though the difference is probably not enough to justify its higher price. The biggest and easiest SQ jumps are still to be had by cutting the unused pairs and by running the LAN as slowly as possible (in my case, 10-BaseT).

 

Meanwhile, in CAT and pigeons mode, I also recently (for non-audio reasons) swapped a ZyXel ES105a 10/100 (version 3) switch sited just before my Fit-PC2 "endpoint" for version 1 of the box . Intrigued by how the older device sounded notably better, I saw that it uses an older chip though whether that's the cause I can't say. (My digital setup is all LPS-1 powered.)

[Ryelands]

Has anyone done a  starquad DC > SATA power cable . . .

 

Yes. I use an LPS-1 set to 5v to power an SSD, an Intona and a USB>I2S card. I hand-wound a wiring harness from 24 awg cables with quad-core geometry and shielded it with self-adhesive copper tape wired JSSG-style. I used a scrap  connector for the SSD. (As a precaution, I did try separately powering the SSD in case sharing the PSU affected the sound quality of the other devices but it didn't seem to.)

 

I also shielded the SSD and its data lead with self-adhesive copper sheet and configured them as another Faraday shield. I can't really say how much difference that made as I've now configured every digital device in my system as Faraday shields and, for obvious reasons, did so to several at a time. What I can say is that the improvement overall was dramatic and well worth the effort.

 

HTH

 

[Ryelands]

On 27/01/2018 at 7:52 PM, JohnSwenson said:

. . . going with the starquad is probably better than the twisted pairs.

 

I've been using two-core Supra speaker cables for several years. In light of discussions here and elsewhere and as Supra describes quad-core as "probably the best speaker cable we make", I thought I'd try it.

 

Then I saw the price. Eventually, however, I DIY'd by buying more two-core cable and making up the QC runs manually, i.e. by "upgrading" the old cable. I wrapped the runs with PTFE tape to keep them tightly twisted, fitted a JSSG-style braided-copper sleeve (mostly for mechancial protection) and finished off with the usual poly-whatever braided sleeve.

 

If you want to learn what wrestling with a bad-tempered cobra is like, the procedure will give you a fair idea BUT the difference between the two-core and the quad-core runs was very marked. True, the resistance of the new cable is half that of the old but trials on a different system suggest that the improvement is mostly, if not entirely, down to the altered geometry. (Ditto for power cables BTW but that's a story for another time.)

 

So many thanks to John for the idea which I can confirm works very well.

 

Dave